Dichlorophenol Exposure and the Development of Allergy

This article appeared in the December issue of the Annals of Allergy, Asthma, and Immunology  and has gotten a significant amount of attention. I was interviewed by a local radio station about the article. When you interview for a newspaper or for  radio, you are never sure what the end-product will be. The interviewee is never given the chance to review the final report. All too often comments are taken out of context and you wind up being misquoted. There is also the longing for sensationalism. The reporters do not pay attention to what is important. They have their agenda and already know what their conclusions will be. A dear friend, Wes Burks, MD- President of the American Academy of Allergy, Asthma, and Immunology sent me a note after he read my comments. To my surprise, I thought this was for a local audience. My lesson learned is to only say what you want to see published or reported. I was very tempted to put a halt to these interview requests. When the requests come in, they want the information yesterday and they do not forward the article that is under review. This forum allows me to make it right.

I have a format for reviewing articles. I learned this from Dr. Hopple- Professor of Pharmacology at Case Western Reserve University during my graduate student years. The art of review remains with me. First you look at the journal and decide if it is a high tier publication or truly a magazine. Then you look to the group, the sponsorship, and the duration from submission to acceptance. Next look for the question that is asked in the paper or what was the purpose of the work. The results will only be as good as the methods used to derive them. Make sure the techniques used are valid and can answer the purpose. Look at the charts and tables, draw your own conclusions then see if the discourse regarding results matches your interpretation. Compare your conclusions with those of the author. The exercise is fun and often you see where there may be a problem or you will see what the author was really trying to say. Just don’t read the abstract- analyze the paper.

The focus here will be on food allergy part of the study. The study found no difference with environmental allergies and dichlorophenol (DCP) exposure.

The journal is very good. The group is excellent- Dr. Rosenstreich is the senior author and he is very well respected. The timeline from submission to acceptance was five months- not too bad.

The purpose of the paper: To examine the relationship between exposure to chlorinated phenols in a national representative sample of the general United States population and the prevalence of sensitization to foods and environmental allergens.

Methods: the sample came from the National Health and Nutrition Examination Survey (NHANES) 2005-2006. Urine samples were evaluated for chlorinated phenols. Those who had both urine samples and blood tests for allergy were included.  No one in the study was less than 6 years of age.

Pesticide exposure was determined by a questionnaire and by measuring compounds in the urine. Allergen sensitization was determined by a blood test. The foods that were included were;

  • Peanut
  • Egg
  • Milk
  • Shrimp

Sensitization and food allergy was defined as any detectable level of specific IgE directed against the food.

Results: The first table shows the characteristics of the population studied. There were 2,211 participants.

            There were 411 (18.6%) who had a positive IgE to one of the foods tested. The average age was 39.5 in the population. Most (71.1%) were non-Hispanic white. The dichlorophenols levels were listed as low, high for one, and high for two. A high value was defined as being > 75% tile. There was a significant difference between food sensitized groups and dichlorophenol levels with more allergic individuals having higher levels. There was more asthma and nasal allergy in the food-sensitized group. The table shows 6% of those with sensitization to food had used pesticides in the home to control insects compared to 10% of the group with no food sensitization. This difference was significant. Urban living also was found to be significantly associated with food sensitization.

            The only figure in the paper was a graphic representation of dichlorophenol levels and the percent with had food allergy; there was a significant difference in the prevalence of food allergy between the low and one or two high urinary dichlorophenol groups.

  • Low level- ~13% had food allergy (of 1,555)
  • One high level-~20% had food allergy (of 225)
  • Two high levels-~23% had food allergy (of 431)

            The second table in the article shows the odds ratios (ORs). These come from a multivariate analysis of the information. The significant findings were;

  • Male sex    OR = 1.8
  • Non-Hispanic Black       OR = 1.8
  • One elevated DCP          OR = 1.5
  • Two elevated DCPs        OR = 1.8
  • Low vitamin D level       OR = 1.7
  • Presence of asthma       OR = 1.6
  • Presence of Hay Fever  OR = 2.5
  • Pesticide in Home         OR = 0.5

(an odds ratio of 1.8 means that the risk of having the condition- allergy is 1.8 times higher. Here, if you have hay fever, the odds that you have a food sensitization is 2.5 times higher than someone without hay fever.)

Conclusions: High dichlorophenol level in the urine was significantly associated with sensitization to food. This is seen without environmental allergy.

Discussion (from the paper)

            First study so show associated of dichlorophenols and allergic sensitization

            Relevance questioned- food allergy seen more in children. Could this be a marker for the adult?

            The age at sensitization is not known.

            Exposure possible via chlorinated water or food treated with the compounds.

            Sensitization to foods in adults is increasing.

            The study’s purpose was not to determine the mechanism of the sensitization.

            Exposure to dichlorophenol pesticides could weaken tolerance to foods.

            Dichlorophenols may weaken the immune system.

            Widespread use of bottled water offsets impact of exposure from chlorination.

            Study limitations

                        Cause and effect not established- this was a cross-sectional study.

                        No inclusion of young children where food allergy is the greatest.

                        Only 0.7% had IgE levels at a high predictive value for a reaction.

Reviewer’s comments:

            First, this was a very interesting paper. I enjoyed reading it and it did for me what a good report should do, it made me ask more questions. This is a valuable contribution. It clearly demonstrates an association. The next step is to go further with investigations regarding any cause/effect issues.

            Now for a few concerns, those things you pick up when you look at every word and table/figure.

            The authors go back and forth with two terms that you need to be careful with. What they have looked at here was sensitization. They looked at an antibody level to a limited number of foods. There is nothing in the paper that tells us that the population was allergic. Allergy would be what we really want to know and allergy is that unique combination of a history of a reaction with exposure and a positive test to support it. We have half of the criteria, yet the text and table clearly say allergy.

            We know that any laboratory test for a food allergy has a significant false positive rate. The authors tried to address the issue of allergy by looking only at IgE levels that are known to be highly predictive of a reaction and only had 0.7% of the individuals fall within that range. Those 29 patients would be the ones that had the higher risk of an allergic reaction. So we really do not know if true food allergy existed. We do see sensitization only.

            The authors pointed this out – the population that was evaluated was not where food allergy is so much more common- our children. The findings are limited to a population who on the average are 39.5 years of age.

            It would have been great to see the food allergy history.  Two of the food sensitivities, egg and milk, can be outgrown in 80%. Peanut allergy can resolve in 20%. Shrimp allergy may be life-long. The paper did not show the breakdown of who was sensitive to what.

            A point of confusion also occurred- insecticide use in the homes of the food sensitized group was 6% and 10% of home had these products in the group without food sensitization. So there is less use and the odds ratio was 0.5. So is there not less use of pesticides (6 vs.10% in the sensitized) and an OR < 1.0 which shows less risk (protection?).

            The figure in the paper is a nice representation of dichlorophenol levels and the percent of food allergy related to those levels. Here the author used allergy where in the table it is sensitization to food. I tried to do the math and I can’t match what I see in the figure. What I came up with is

            Food ‘Allergy’

  • Low DCP level                                 16%
  • One elevated DCP compound        27%
  • Two elevated DCP compounds      29%

We have had chlorination of the water since 1908. Chlorine gas has been used since 1910. Water-borne infections have decreased significantly with a clean water supply. The exposure to chlorine by-products has been out there for years. It is possible that there is more consumption- with all the fears of dehydration. The adult tends to favor bottled water. So the exposure is more from other, non-tap water sources.

What also needs to be considered is could this be an observation due to the behavior of people with food allergy? An epi-phenomenon where we see high levels because of a food allergy problem

This is very interesting. This is epidemiology and truly food allergy/sensitization is a public health problem.

FEL 12-12-12

December 12, 2012 · fleickly · No Comments
Tags: , , ,  · Posted in: Allergies, Article Review, Developing Allergy, Environment, Food Allergies, Interesting articles